My cover is blown

“Are you a spy?” is a question I get surprisingly often.Yes, I was raised in the Soviet Union and have spent much of my life making relatively little sense to anybody anywhere. Still, the question surprises me mostly because regular people, tickled by the possibility that they are standing next to a maybe-spy, seem to expect a meaningful answer.If I were a spy (and I never confirm nor deny), I would have been fielding that question since I first learned to regard a pacifier with suspicion. Also, if I were a spy, I wouldn’t have any accent at all. Haven’t people seen The Americans?Anyway, with AI taking over the world, I can no longer hide the truth.Spotify has taken in hundreds of hours of my choices across music, podcasts, audiobooks and, yes, recommendations from friends — really, really watch who you allow to contaminate the evidence — and noticed something many of you have treated as a running joke for years.I have spent much of this summer noticing.Since I am currently based in North America, summer is only now coming to a close, leaving me up to my neck in unprocessed evidence.I have spent most of my professional life observing people, organizations, systems, movements: what changes, what stubbornly doesn’t, and what happens between intention and reality. Observation is practically an occupational hazard.But my PhD has been forcing me to distinguish more carefully between looking in order to reach a conclusion and looking to acknowledge what is there. Age may also be playing a role. Not chronological age, necessarily. I don’t remember who said it first, but I was with a group of friends when it was established that I was 76 on the inside. That was at least ten years ago. When I got perilously close to an internal 90, I decided something had to change or I’d soon be the oldest person anybody’s ever met. Lonely.I have been attempting a “Benjamin Button”: methodically reverse-aging my inside person.I reluctantly accepted that mac-and-cheese is legitimate food. Tried and was smitten by a weighted blanket. Gradually—over months—crept my bedtime from 8:15 p.m. toward 10. Committed to phone puzzles. Got hair tinsel. Glitter. A dog bed. All for myself.Impressive, I thought.And then yesterday afternoon, when asked for my Daylist, Spotify analyzed me and labeled it, coastal grandmother.Correct. Coastal grandmother.See Exhibit A.I sent this screenshot to several friends because, obviously, an accusation of this magnitude requires sufficiently appalled witnesses. This only becomes a story because my friends weren’t appalled. One friend reminded me that I had been claiming to be 74 for years. I corrected him. It has never dropped below 76 in recorded history, so there was no need to start flattering me now.What disturbed me was not that Spotify had got me wrong. It was that Spotify had somehow got me right.The algorithm looked past much of the evidence humans tend to use when deciding who somebody is and landed on something I have been insisting was true all along. Not even especially deep down.The evidence against me“Coastal grandmother” — I had to know, in case it comes up in New York Times Puzzles; thank you, Glamour — was coined and popularized in 2022 by TikTok creator Lex Nicoleta. You do not have to be a grandmother or live anywhere near a coast: this type of person is apparently excused from all integrity. The archetype draws heavily on Nancy Meyers heroines: relaxed linen, good food, farmers’ markets, fresh flowers, understated elegance, early mornings spent holding artisanal mugs with both hands, and effortless enjoyment of one’s life while expressing unfounded doubt and disorientation. Architectural Digest describes the domestic version in terms of natural materials, soft colors, eclectically washed out homes swaddled in light, comfort, simplicity, and everyday pleasure.In my mind, she has also perfected the self-deprecating chuckle and the lollop: that oddly loose, over-articulated gait, with too much limb, too much side-to-side, and not enough obvious reason.S**t.Diane Keaton has been preparing me for this diagnosis for decades.Once you scrub away the baby face, the midlife chub, my actual date of birth and other inconvenient forensic evidence, Spotify may have produced a surprisingly accurate profile. Once I stopped laughing, I became interested in the data Spotify had apparently found beneath a running joke.Caricatures work because they are wrongA good caricature is intentionally inaccurate. The proportions are all wrong. The eyebrows have become an independent life-form. Something recognizable has been exaggerated almost to the point of insult. And still, the person may be more recognizable than in a photograph.The distortion works because somebody has correctly identified the features carrying an unusual amount of information. This may be one useful way to think about both artificial and human intelligence.We are constantly making little models of one another.She is ambitious.He is shy.They are generous.He is witty.She is difficult.He is unreliable.They are eccentric.Eventually: Ah. This is you.Except every one of those conclusions can be uncannily accurate and radically incomplete at exactly the same time. We have been doing this to one another long before machines joined in. Perhaps what AI changes is the scale of the evidence.The interesting question is not whether the machine has captured my identity. Of course it hasn’t. The interesting question is: what did it notice that other people discounted because it contradicted more obvious evidence?I don’t look or lollop 76. I actually feel younger than I have in decades. Spotify does not know me. I merely left such a pervasive trail pointing to rayon and chai that the date of birth on my profile presented an anomaly.What counts as evidence of a person?This question interests me professionally because much of my work as a transformation expert involves pattern recognition.I am trying to understand changemakers well enough that we can support their vital function in the world better than we currently do. That requires categories. Are there recurring ways people initiate change? Build it? Legitimize it? Hold relationships and spaces around it? Steward what has been created? Almost certainly. But the moment we identify a pattern, another question arrives: When is categorization clarifying, and when does it flatten a human being?I don’t think the answer is to stop recognizing patterns. That would be absurd. Pattern recognition is one of the ways we know anything at all. The harder discipline may be learning to let a pattern tell us: Look here. How curious. Not: Gotcha! Case closed.That feels important far beyond my research.What criteria do we use to decide what most deeply defines another person? Age? Profession? Appearance? Diagnosis? Politics? Income? The worst thing they have done? The best? What they are frightened of? What reliably delights them? What they notice? What they protect? What they return to? Under what particular conditions do they become most alive? I am increasingly interested in those last kinds of evidence.Maybe individuality lives less in our demographic facts than in our peculiar patterns of sensibility, sensitivity and delight-ability.The things that move us.The things that reliably commandeer our attention.The things we cannot quite help loving.Better caricatures. Not flatter people.However delightful — and disturbing — it was to be validated as a coastal grandmother, I don’t want technology that tells me who somebody is. Humans have already demonstrated how dangerous we are when convinced we have other people figured out. But I can imagine technology becoming extraordinarily useful at helping us notice.You seem happier after you speak with this person.You keep returning to this kind of music.You say this matters to you, but your effort is elsewhere.You may be more this than you thought.Imagine computational pattern recognition paired not with human certainty, but human curiosity.Not, The data says this is you. Rather, Something keeps appearing here. Shall we look?That seems to me a much more interesting place for artificial intelligence and human knowing to meet. The machine gets better at recognizing patterns. We get better at refusing to conflate a pattern with the whole person. And perhaps, between those capabilities, we become a little better at seeing one another for what is already true —and for everything that might yet become possible from it.In the meantime, Spotify has apparently concluded that beneath my current identity experiments lives a woman who savors music, good food, plants, soft fabrics, long walks to meet herself, people to love, and enough quiet to notice her own life.My cover is blown.I can live with that._ _ _Changemakers’ Handbook is an audience-supported publication focused on professionalizing chaangemaking in a post-solutions world. Consider subscribing to invest in this work, join future live conversations and access all posts. This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit changemakershandbook.substack.com/subscribe